Study On Outstanding Claims Reserve Of R&C Based On Generalized Addictiv Mixed Model

With the development of the insurance industry in our country, the competition of insurance industry becomes fiercer and fiercer which to aggravate the risk day by day.because of the feature of insurance,the risk is reflected in the level of the Claims Reserve mainly. Accurate estimation of the level of the outstanding Claims Reserve for an insurance company is required because it is the major liability in insurance company's balance sheet and it impacts on multiple aspects of the operation of the insurance company: dividend declaration,solvency,insurance product pricing and tax payments.Chain ladder method has been widely used in loss reserving,it is very simple,and many insurers prefer the method.But it also has many disadvantages.For business with long tail or new business,when the payment data is insufficient,we can not get the payment patterns of the tail, The reason that it has been restrict for using in the case of former root from its static.then people try to introduce the stochastic model to the reserving.The Generalized Linear Model has been used widely in the these stochastic models,but it has the disadvantages for ignoring the non-linear effect of covariable and stochastic from exterior.we introduce the Generalized Addictive Mixed Model to the reserving,then construct the Gamma- Addictiv Mixed Model and Poisson Addictiv Mixed Model and conbine bayse and MCMC method,then get the result of reserving.Comparing with traditional chain ladder method,the GAMM is more exact for the long tail.Also,we can get the distribution of reserves and the payment patterns of every accident years,so it provides much information about loss reserving and insurer's on-going operation.